To determine the effects of the US economy on heart failure hospitalization rates.
To determine the effects of the US economy on heart failure hospitalization rates.
The recession was associated with worsening unemployment, loss of private insurance and prescription medication benefits, medication nonadherence, and ultimately increased rates of hospitalization for heart failure.
We compared hospitalization rates at a large, single, academic medical center from July 1, 2006 to February 28, 2007, a time of economic stability, and July 1, 2008 to February 28, 2009, a time of economic recession in the United States.
Significantly fewer patients had private medical insurance during the economic recession than during the control period (36.5% vs 46%; P = 0.04). Despite this, there were no differences in the heart failure hospitalization or readmission rates, length of hospitalization, need for admission to an intensive care unit, in-hospital mortality, or use of guideline-recommended heart failure medications between the 2 study periods.
We conclude that despite significant effects on medical insurance coverage, rates of heart failure hospitalization at our institution were not significantly affected by the recession. Additional large-scale population-based research is needed to better understand the effects of fluctuations in the US economy on heart failure hospitalization rates. Clin. Cardiol. 2012 DOI: 10.1002/clc.21996
The authors have no funding, financial relationships, or conflicts of interest to disclose.
Heart failure affects 5.7 million people, has an incidence of 670 000 in those over the age of 45 years, and causes 56 565 deaths in the United States anually.1 This disease burden leads to 990 000 hospitalizations for heart failure each year and results in an estimated $20.9 billion expenditure in direct hospitalization costs annually.1,2 Although there has been much effort to reduce heart failure hospitalizations and improve quality of care in heart failure, hospitalization rates remain problematic, and rehospitalization after discharge approaches 50% at 6 months.3 Furthermore, the factors that promote these high rates of hospitalization remain poorly understood, and trials that have attempted to identify predictors of admission for heart failure have focused primarily on patient-specific variables including disease severity, comorbidities, laboratory values, and medication/dietary compliance.4,5 Little is known about the effects of economic variables on heart failure readmission rates.
A variety of studies aimed at determining the cause of heart failure readmission have identified medication nonadherence as the most common predicating factor for admission, and medication noncompliance may contribute to as many as 10% of all heart failure hospitalizations.5–7 Several economic factors, such as poor financial status, lack of medical insurance, and inadequate prescription benefits, have been implicated as underlying causes of medication nonadherence. In the Get With The Guidelines–Heart Failure registry, which included 54 322 heart failure patients, nonadherent patients were more likely to be uninsured than adherent patients.5 Poor financial status, as defined as not having enough money to make ends meet, has also been associated with medication nonadherance in a multivariate analysis of 134 heart failure patients.6 Several studies have reported an association between inadequate prescription medication benefits and medication noncompliance. In 1 trial evaluating Medicaid beneficiaries, 42% of patients failed to fill medical prescriptions when costs exceed the limits of the Medicaid prescription cap.8 A separate study, which included 3073 patients over the age of 67 years with at least 1 chronic medical condition, found that those without a prescription benefit acquired fewer medication classes and experienced lower refill adherence than those with a prescription benefit.9 A similar evaluation of 199 179 Medicare beneficiaries reported that patients whose annual drug benefits were capped took fewer of their prescribed medications and had more emergency room visits and hospitalizations (relative risk [RR]: 1.09 and 1.13, respectively) than patients with unlimited annual drug benefits.10
In several population-based studies, unemployment has been associated with adverse health outcomes including increased overall mortality and mortality due to cardiovascular disease and suicide.11 For example, in a census-based study of Finish wage-earning men age 30 to 54 years covering nearly 2.7 million person-years, unemployed individuals had increased rates of death due to cardiovascular disease (RR: 1.54; 95% confidence interval: 1.40–1.70), and the effects of unemployment on mortality were more pronounced with increasing duration of unemployment.12 Unemployment has also been associated with increased use of health services, and unemployed individuals have been reported to be between 0.2 to 5 times more likely to visit physicians and 0.33 to 2 times more likely to be admitted to the hospital then their employed counterparts.11,13–15
Despite this understanding, the relationship between economic forces and health outcomes is poorly understood. The purpose of this study was to evaluate the effects of the recent economic recession in the United States on heart failure hospitalizations. We hypothesized that the recession was associated with worsening unemployment, loss of private insurance and prescription medication benefits, medication nonadherence, and ultimately increased rates of hospitalization for heart failure.
We performed a retrospective search through hospital billing records from a large academic medical center to identify patients hospitalized for heart failure from July 1, 2006 to February 28, 2007 (7 months), a time of economic stability (control group), and from July 1, 2008 to February 28, 2009 (7 months), a time of economic recession in the United States as evidenced by declines in the Dow Jones Industrial Average (recession group) (Figure 1). All patients admitted with a primary diagnosis of decompensated heart failure with a known history of chronic advanced heart failure were included in the study. The diagnoses were verified by combining the International Classification of Diseases, 9th Revision billing codes (Table 1) with the actual patient charts' admission diagnoses and discharge diagnoses. Patients with suspected heart failure at admission but with a different main diagnosis at discharge (eg, paroxysmal a trial fibrillation leading to heart failure, pneumonia, angina pectoris leading to heart failure) were also excluded to detect only primary cardiac decompensation on top of a chronic known condition independent of other external or internal conditions. Patients with a history of heart transplantation or implantation of a ventricular assist device were also excluded from the study. Patients with Medicare were also excluded as their access to medical care was thought to be more dependent on Medicare eligibility and policy than on economic forces. Patients with newly diagnosed heart failure were also not included.
|402.01||Malignant HTN with HF|
|402.11||HTN with HF|
|402.91||HTN heart disease, unspecified, with HF|
|404.01||HTN heart and kidney disease, malignant|
|404.03||HTN heart and kidney disease, malignant|
|404.11||HTN heart and kidney disease, benign|
|404.13||HTN heart and kidney disease, benign|
|404.91||HTN heart and kidney disease, unspecified|
|404.93||HTN heart and kidney disease, unspecified|
|428.0||Congestive HF, unspecified|
|428.20||Systolic HF, unspecified|
|428.21||Systolic HF, acute|
|428.22||Systolic HF, chronic|
|428.23||Systolic HF, acute on chronic|
|428.30||Diastolic HF, unspecified|
|428.31||Diastolic HF, acute|
|428.32||Diastolic HF, chronic|
|428.33||Diastolic HF, acute on chronic|
|428.40||Combined systolic anddiastolic HF, unspecified|
|428.41||Combined systolic anddiastolic HF, acute|
|428.42||Combined systolic anddiastolic HF, chronic|
|428.43||Combined systolic anddiastolic HF, acute on chronic|
Chart review allowed for the documentation of patient baseline demographics, length of hospitalization, home medication regimen, medical insurance provider, need for intensive care unit (ICU) admission and length of stay, and in-hospital mortality. Between groups, comparisons were achieved with 2-tailed equal variance t tests; within-group comparison was achieved using 2-tailed paired t tests. A P value of <0.05 was considered statistically significant.
There were 141 admissions for acutely decompensated heart failure during the control period and 137 admissions during the recession. These admissions were accounted for by 113 patients with a 20% readmission rate and 104 patients with a 24% readmission rate in each group, respectively (all P = not significant) (Table 2). There were no between-group differences in patient baseline demographics, length of hospitalization, need for ICU admission, in-hospital death, or use of guideline-recommended heart failure medications. Significantly fewer patients had private medical insurance during the economic recession than during the control period (36.5% vs 46%; P = 0.04). There were significantly more admissions for decompensated heart failure secondary to diastolic heart failure during the control period than during the recession (38% vs 25%, P = 0.03), which resulted in differences in mean ejection fractions between groups (40% vs 33%, P = 0.004).
|Recession, No. (%), n = 137||Control, No. (%), n = 141||P Value|
|HF readmissions||33 (24.1)||28 (19.9)||NS|
|No. of patients||104||113||NS|
|Male||71 (68.3)||73 (64.6)||NS|
|Female||33 (31.7)||40 (35.4)||NS|
|Ejection fraction, average||33||40||0.004|
|Systolic HFa||72 (69.2)||66 (58.4)||NS|
|ACE/ARB||45 (62.5)||34 (51.5)||NS|
|β-blocker||42 (58.3)||34 (51.5)||NS|
|Aldactone||13 (18.0)||11 (16.7)||NS|
|Nitrate||7 (9.7)||8 (12.1)||NS|
|Hydralazine||4 (5.6)||2 (3.0)||NS|
|Digoxin||13 (18.0)||20 (30.3)||NS|
|Statin||22 (30.6)||17 (25.8)||NS|
|Loopdiuretic||43 (59.7)||42 (63.6)||NS|
|Diastolic HFb||26 (25)||43 (38.0)||0.03|
|ACE/ARB||7 (26.9)||22 (51.2)||0.02|
|β-blocker||13 (50.0)||26 (60.5)||NS|
|Aldactone||0 (0.0)||5 (11.6)||0.01|
|Nitrate||6 (23.1)||4 (9.3)||NS|
|Hydralazine||5 (19.2)||3 (7.0)||NS|
|Digoxin||1 (3.8)||7 (16.3)||0.04|
|Statin||7 (26.9)||12 (27.9)||NS|
|Loopdiuretic||9 (34.6)||22 (50.2)||NS|
|Length of stay, average, d||6||7||NS|
|ICU admissions||37 (27.0)||32 (22.7)||NS|
|Length of ICU stay, average, d||3||6||NS|
|Private medical insurance||38 (36.5)||52 (46.0)||0.04|
|Medi-Cal/Medi-Cal managed care||60 (57.7)||54 (47.9)||NS|
|Other funding||7 (6.7)||7 (6.2)||NS|
The economic recession in the United States was associated with a significant reduction in the number of patients with private medical insurance at our hospital; however, it did not affect the frequency of hospitalization, readmission rates, length of hospital stay, need for admission to an ICU, in-hospital mortality, or use of heart failure medications in patients with acute decompensated heart failure. The 9.5% reduction in the number of patients with private medical insurance observed in this study is similar, albeit more pronounced, than national trends in the rates of employment-based medical coverage, which were 61.3% in May 2007 and 58.2% in July 2009.16 The reasons for these differences are unclear, but may be related to differences between regional and national unemployment rates. For example, the national uninsured rate increased by 4.1% between 2007 and 2009, whereas it increased by 28% in California during the same time period.16,17 This suggests that our patient population may have included a larger number of uninsured patients than the general US population, and that our population may have been at a higher risk of economic influences. That our study failed to identify an association between the recession and hospitalization rates in this population adds strength to our findings.
This study has several limitations including its retrospective nature, use of single-center data, and limited sample size. Additionally, individual patient factors, which would be expected to influence our results, such as employment status, income, use of government assistance programs such as Welfare, and medication compliance, were unavailable.
The relationship between fluctuations in the US economy on health outcomes is poorly defined. Our study is significant because it is the first to evaluate economic effects of the recession on heart failure outcomes. Available data suggest that economic factors such as unemployment, financial well-being, and insurance coverage are associated with medication compliance and use of health resources. Given this, we expected that the recession would lead to unemployment, loss of medical insurance, poor medication compliance, and increased hospitalization rates; however, this effect was not demonstrated. These results are surprising given the significant reduction in the number of patients with private medical insurance admitted to our institution during the recession. The reasons for our findings are unclear and may be due to a variety of factors. First, it is possible that government health coverage programs were able to compensate for the loss of private insurance during the recession. Although nonsignificant, our data did demonstrate a 6% increase in the number of patients admitted with Medi-Cal (a California-based Medicaid program) coverage during the recession. Second, heart failure hospitalizations may be primarily driven by noneconomic factors, such as severity of disease, inadequacy of current heart failure therapy, poor dietary and lifestyle choices, and comorbid illnesses, and that frequent hospitalization remains problematic regardless of the status of the economy. Third, is the possibility of the inverse-care law, the theory that uninsured patients fail to seek medical attention due to concerns about the inability to pay for medical services18; it is possible that the recession led to reduced hospitalizations in uninsured patients. Uninsured patients who did not seek medical attention for their heart failure were not included in our study, which could have affected our results. Finally, our single-center experience may not accurately represent regional or national trends in hospitalization rates, which limits the strength of our findings.
We conclude that heart failure hospitalization rates may not be as susceptible to changes in economic forces as predicted, and that heart failure hospitalization rates have remain suboptimal even during times of economic prosperity. Additional large-scale population-based research is needed to better understand the effects of economic influences on heart failure hospitalization rates.